1521-0103/367/2/194–202$35.00 https://doi.org/10.1124/jpet.118.249581 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 367:194–202, November 2018 Copyright ª 2018 by The American Society for Pharmacology and Experimental Therapeutics

Impairment of Nitric Oxide Pathway by Intravascular Plays a Major Role in Mice Esophageal Hypercontractility: Reversion by Soluble Guanylyl Cyclase Stimulator

Fabio Henrique Silva, Kleber Yotsumoto Fertrin, Eduardo Costa Alexandre, Fabiano Beraldi Calmasini, Carla Fernanda Franco-Penteado, and Fernando Ferreira Costa Hematology and Hemotherapy Center (F.H.S., K.Y.F., C.F.F.-P., F.F.C.) and Department of Pharmacology, Faculty of Medical Sciences (E.C.A., F.B.C.), University of Campinas, Campinas, São Paulo, Brazil; and Division of Hematology, University of Washington, Seattle, Washington (K.Y.F.) Downloaded from Received April 1, 2018; accepted July 30, 2018

ABSTRACT Paroxysmal nocturnal hemoglobinuria (PNH) patients display cyclase stimulator 3-(4-amino-5-cyclopropylpyrimidin-2-yl)- exaggerated intravascular hemolysis and esophageal disor- 1-(2-fluorobenzyl)-1H-pyrazolo[3,4-b]pyridine (BAY 41-2272; ders. Since excess hemoglobin in the plasma causes re- 1 mM) completely reversed the increased contractile responses jpet.aspetjournals.org duced nitric oxide (NO) bioavailability and oxidative stress, we to CCh, KCl, and EFS in PHZ mice, but responses remained hypothesized that esophageal contraction may be impaired unchanged with prior treatment with NO donor sodium nitro- by intravascular hemolysis. This study aimed to analyze the prusside (300 mM). Protein expression of 3-nitrotyrosine and alterations of the contractile mechanisms in a 4-hydroxynonenal increased in esophagi from PHZ mice, sug- murine model of exaggerated intravascular hemolysis induced gesting a state of oxidative stress. In endothelial nitric oxide by phenylhydrazine (PHZ). For comparative purposes, sickle synthase gene-deficient mice, the contractile responses elicited cell disease (SCD) mice were also studied, a less severe by KCl and CCh were increased in the esophagus but remained intravascular hemolysis model. Esophagus rings were dis- unchanged with the intravascular hemolysis induced by PHZ. In at ASPET Journals on September 30, 2021 sected free and placed in baths. Plasma hemoglobin conclusion, our results show that esophagus hypercontractile was higher in PHZ compared with SCD mice, as expected. The state occurs in association with lower NO bioavailability due to contractile responses produced by carbachol (CCh), KCl, and exaggerated hemolysis intravascular and oxidative stress. electrical-field stimulation (EFS) were superior in PHZ esophagi Moreover, our study supports the hypothesis that esophageal compared with control but remained unchanged in SCD mice. disorders in PNH patients are secondary to intravascular hemo- Preincubation with the NO-independent soluble guanylate lysis affecting the NO-cGMP pathway.

Introduction The consists of a muscular tube that extends from the mouth to the anus. The esophagus is part of – The nitric oxide (NO) cGMP signaling pathway is well the gastrointestinal tract, and its function is transporting known to regulate smooth muscle relaxation and contraction swallowed food into the and preventing gastroesoph- (Förstermann and Sessa, 2012). Upon stimulation by NO, ageal reflux (Yazaki and Sifrim, 2012). Esophagus smooth the smooth muscle soluble guanylate cyclase (sGC) enzyme muscle receives a rich cholinergic excitatory innervation, catalyzes the production of cGMP following activation of resulting in the discharge of that acts on the cGMP-dependent protein kinase, which leads to smooth muscarinic receptor of smooth muscle and promoting esoph- muscle relaxation from different kinds of tissues (Francis ageal smooth muscle contractions (Goyal and Chaudhury, et al., 2010; Förstermann and Sessa, 2012). Alterations of 2008; Braverman et al., 2009). NO produced by nitrergic the NO signaling pathway in the vascular or nonvascular is considered the principal inhibitory neurotransmitter smooth muscle have been associated with several disorders, that promotes smooth muscle relaxation of the esophagus such as cardiovascular disease and gastrointestinal disease (Shah et al., 2004; Goyal and Chaudhury, 2008). A previous (Mearin et al., 1993; Förstermann and Sessa, 2012). study showed mRNA expression of endothelial nitric oxide synthase in human esophageal mucosa (Inamori et al., 2006). Esophagus contraction induced by cholinergic stimula- F.H.S. and F.F.C. thank São Paulo Research Foundation (FAPESP) for financial support [Grants 2014/00984-3, 2017/08122-9, and 2018/06243-6]. tion is increased by NO production inhibition, as well as https://doi.org/10.1124/jpet.118.249581. esophageal peristalsis (Yamato et al., 1992; Lecea et al., 2011).

ABBREVIATIONS: BAY 41-2272, 3-(4-amino-5-cyclopropylpyrimidin-2-yl)-1-(2-fluorobenzyl)-1H-pyrazolo[3,4-b]pyridine; CCh, carbachol; EFS, electrical-field stimulation; eNOS2/2, endothelial nitric oxide synthase gene-deficient; 4-HNE, 4-hydroxynonenal; NO, nitric oxide; 3-NT, 2 3-nitrotyrosine; O2 , superoxide anion; pEC50, potency; PHZ, phenylhydrazine; PNH, paroxysmal nocturnal hemoglobinuria; SCD, sickle cell disease; sGC, soluble guanylate cyclase; SNP, sodium nitroprusside.

194 Effects of Intravascular Hemolysis on Esophagus 195

A previous study reported that guanylyl cyclase gene–deficient We injected PHZ at 50 mg/kg into C57BL/6 and eNOS2/2 mice mice display higher esophagus contractions compared with wild intraperitoneally to induce intravascular hemolysis. Mice were rein- mice (Groneberg et al., 2015). jected with 50 mg/kg 8 hours later (Lim et al., 1998) and were killed Paroxysmal nocturnal hemoglobinuria (PNH) is a rare with inhaled isoflurane 4 days after the end of PHZ treatment. acquired clonal hematopoietic stem cell disease in which C57BL/6 and PHZ mice received deferoxamine (150 mg/kg per day via i.p. injection, from second to fourth day), deferiprone (50 mg/kg per uncontrolled complement activation leads to intravascular day, given as daily gavage from second to fourth day), or vehicles hemolysis (Oni et al., 1970; Hill et al., 2017). PNH patients (saline and water, respectively). display /odynophagia, and these alterations seem to Hematologic Parameters. Whole blood was obtained by intra- be due to esophageal spasm, but the mechanisms underlying cardiac puncture from ketamine/xylazine-anesthetized mice and anti- esophageal dysfunction are largely unknown (Moyo et al., coagulated with ethylenediamine tetraacetic acid for blood count 2004; Parker et al., 2005). Sickle cell disease (SCD) is also a determination within 30–60 minutes of blood collection on Sysmex blood disorder that manifests with hemolytic anemia, but to XN-3000 equipment (Sysmex, Kobe, Japan). the best of our knowledge, SCD patients do not display Lactate Dehydrogenase and Plasma Hemoglobin Measure- esophageal disease. In PNH and SCD, intravascular hemoly- ment. Mice plasma samples were used to quantify the lactate dehy- sis is a common hallmark, but this clinical evidence is less drogenase and hemoglobin through colorimetric assays using the severe in SCD than in PNH (Reiter et al., 2002; Hill et al., Hemoglobin Colorimetric Assay (Cayman Chemical, Ann Arbor, Michigan) and Bilirubin Assay Kit (Sigma-Aldrich, St. Louis, MO) 2010). In intravascular hemolysis, erythrocyte lysis in the ’ according to the manufacturer s procedure. Assays were done in Downloaded from circulation leads to release of free hemoglobin and other duplicate. erythrocyte components into the plasma, such as asymmetric Functional Studies in Esophagus Smooth Muscle Rings and dimethylarginine, arginase-1, and lactate dehydrogenase Concentration-Response Curves. The esophagus was extracted, (Reiter et al., 2002; Hill et al., 2010; Kato et al., 2017). It is dissected, and cut into rings (1.5–2 mm in length). Rings of esophagus well established that NO scavenging by an excess of plasma smooth muscle were placed in a 5-ml organ system filled with Krebs hemoglobin contributes to several clinical complications of solution at 37°C continuously bubbled with a mixture of 95% O2 and intravascular hemolysis (Hill et al., 2010; Kato et al., 2017). 5% CO2 (pH 7.4). Isometric force was analyzed using a Power jpet.aspetjournals.org The contribution of oxidative stress to the deleterious effects Laboratory version 7.2 system (ADInstruments, Sydney, Australia). of intravascular hemolysis has been previously shown (Fibach The esophagus was allowed to equilibrate for 60 minutes under a resting tension of 1 mN. and Dana, 2015). A result of oxidative stress is the inactivation 2 Cumulative concentration-response curves for the muscarinic of NO biologic activity by superoxide anion (O2 ) excess agonist carbachol (CCh; 10 nM to 100 mM) and KCl (1 mM to 1 M) (Pacher et al., 2007). In this study, since cell-free hemoglobin were acquired in esophagus rings from control, PHZ-treated C57BL/6, excess and oxidative stress reduce the amount of bioavailable SCD, eNOS2/2, and PHZ-treated eNOS2/2 mice. Esophagus con- NO, we hypothesized that exaggerated intravascular he- tractile responses to these compounds were acquired in the absence at ASPET Journals on September 30, 2021 molysis could damage normal smooth muscle function in the and presence of NO-independent sGC stimulator 3-(4-amino-5- esophagus. We aimed to evaluate the esophagus smooth cyclopropylpyrimidin-2-yl)-1-(2-fluorobenzyl)-1H-pyrazolo[3,4-b]pyridine muscle contractile function in a murine model that displays (BAY 41-2272; 1 mM, 15 minutes) or NO donor sodium nitroprusside m intravascular hemolysis induced by phenylhydrazine (PHZ). (SNP; 300 M, 15 minutes). PHZ-treated mice have been used to elucidate the intricate Electrical-Field Stimulation in Esophagus Smooth Muscle Rings. Electrical-field stimulation (EFS) was applied to the rings of pathophysiology of hemolysis (Lim et al., 1998; Zhang et al., esophagus smooth muscle placed between two platinum electrodes 2013; Dutra et al., 2014). Hemolysis may be either intravas- connected to a Grass S88 stimulator (Astro-Med, Warwick, RI). EFS cular or extravascular, but there is some controversy over the was conducted at 80 V, 1-ms pulse width, and trains of stimuli lasting amount of intravascular hemolysis in SCD (Crosby, 1955; 5 seconds at varying frequencies. Supra maximum voltage was applied Bunn et al., 2010). There are few studies on this subject, but to assess frequency-response relationships in all preparations using the amount of intravascular hemolysis seems to be less than electrical stimulation. EFS-produced esophagus contractions were 30% of total hemolysis (Crosby, 1955; Bunn et al., 2010). As a completely eliminated by either the muscarinic antagonist atropine comparison, SCD transgenic mice were also studied with the (10 mM) or the voltage-dependent sodium channel blocker tetrodotoxin same procedure. We focused on the effects of intravascular (1 mM), confirming that nerve-caused esophageal contractile re- hemolysis on contractile mechanisms and the NO signaling sponses are caused by acetylcholine discharge. Data Analysis of Functional Assays. Potency (pEC50) was pathway in esophagus smooth muscle. determined by nonlinear regression performed with GraphPad Prism (GraphPad Software, San Diego, CA) with the constraint that F 5 0. All Materials and Methods concentration-response data were fitted to a logistic function expressed by the following equation: E 5 Emax/([1 1 (10c/10x)n] 1 F), where E is Animals and Treatment. Animal procedures and experimental the maximum response induced by agonists; c is the logarithm of the protocols were performed in accordance with the Ethical Principles EC50, the concentration of drug that produces a half-maximal response; in Animal Research adopted by the Brazilian College for Animal x is the logarithm of the concentration of the drug; the exponential Experimentation and followed the Guide for the Care and Use of term, n, is a curve-fitting parameter that defines the slope of the Laboratory Animals (Ministério da ciência, tecnologia, inovações concentration-response line; and F is the response observed in the e comunicações; 2016). absence of added drug. Contractile response data were normalized to

All mice strains were originally purchased from Jackson Laboratories the wet weight of the respective esophagus rings. The Emax is shown in (Bar Harbor, ME). Characterization and breeding were performed at mN/mg wet weight. the Multidisciplinary Center for the Investigation of Biologic Science in Western Blot Analysis. Esophageal tissue was homogenized in Laboratory Animals of the University of Campinas. We used C57BL/6 lysis buffer and centrifuged at 12,000g for 20 minutes at 4°C. male mice (control), endothelial nitric oxide synthase gene-deficient Homogenates containing 50 mg of total protein were run on 4%–20% (eNOS2/2) mice, and male Berkeley transgenic SCD mice, aged Tris-HCl gels (Bio-Rad Laboratories, Hercules, CA) and transferred to 3–4 months, housed three per cage on a 12-hour light–dark cycle. a nitrocellulose membrane. Five-percent nonfat dry milk (Bio-Rad) in 196 Henrique Silva et al.

Tris-buffered saline/Tween was used for 1 hour at 24°C to block esophagi (0.75 6 0.03 mN/mg) in comparison with the nonspecific binding sites. Membranes were incubated for 15–16 hours control group (0.51 6 0.03 mN/mg) (Fig. 2A). We used a KCl at 4°C with the following antibodies: monoclonal anti-3-nitrotyrosine concentration-response curve to assess esophagus contrac- (anti-3-NT; 1:1000, ab7048; Abcam, Cambridge, MA) and polyclonal tions independently of receptor stimulation. Emax values anti-4-hydroxynonenal (anti-4-HNE; 1:1000, ab46545; Abcam). Den- produced by KCl were also higher (P , 0.05) in PHZ esophagi sitometry was analyzed using the ImageJ software (National Insti- (0.45 6 0.01 mN/mg) compared with control mice (0.28 6 tutes of Health, Bethesda, MD). Quantified densitometry results of 3-NT and 4-HNE were normalized to b-actin. 0.02 mN/mg). We found no change in esophagus contractile Drugs. BAY 41-2272 and CCh were acquired from Sigma-Aldrich. responses produced by CCh and KCl in SCD mice (Fig. 2, C Analytical grade was required for all reagents. Either deionized water and D). No significant difference in pEC50 for CCh was or dimethylsulfoxide was used as a solvent, and working solutions found between groups (Table 1). were diluted prior to use. The final concentration of dimethylsulfoxide NO Donor Did Not Modify the Increased Esophagus was 0.1% Contractions. Preincubation of esophagus with NO donor Statistical Analysis. Statistical analysis was performed with SNP (300 mM) did not change the increased esophagus smooth GraphPad Prism (GraphPad Software). Data are expressed as the muscle contractions in PHZ mice induced by CCh (Fig. 3B) or 6 ’ mean S.E.M. of N experiments. Student s unpaired t test and one- KCl (Fig. 3D). No significant changes after SNP preincubation way analysis of variance with Tukey post-test were used when were detected in CCh- and KCl-caused esophagus contractions comparing two or more than two groups, respectively. A P value below of control mice (Fig. 3, A and C, respectively; Table 1).

0.05 was considered statistically significant. Downloaded from sGC Stimulator Reversed the Increased Esophagus Contractions. Preincubation of esophagus with BAY 41-2272 m Results (1 M) completely reversed the greater Emax produced by KCl or CCh in PHZ mouse esophagus to control values (Fig. 4). No Hematologic Parameters. Hematologic parameters (red significant changes after BAY 41-2272 preincubation were blood cells, total hemoglobin, hematocrit, lactate dehydrogenase, detected in CCh- and KCl-induced esophagus contractions of and reticulocyte) showed that PHZ and SCD mice had severe control mice (Fig. 4, A and C, respectively). No significant jpet.aspetjournals.org anemia compared with control mice (Fig. 1). Plasma hemoglobin change in pEC50 for CCh was found between groups (Table 1). was greater in PHZ and SCD compared with the control group Contractile Responses Induced by Cholinergic Nerve (Fig. 1C). Moreover, plasma hemoglobin was higher in PHZ StimulationinEsophagus.EFS produced frequency- compared with SCD mice, indicating that intravascular hemo- dependent contraction esophageal rings (8–32 Hz) in both lysis, as expected, is more exaggerated in PHZ mice than in SCD control and PHZ mice (Fig. 5A). EFS-caused contractions were (Fig. 1C). significantly higher in esophagi from PHZ mice than in control

Exaggerated Intravascular Hemolysis Leads to Esoph- mice at all frequencies tested (Fig. 5A). Preincubation of at ASPET Journals on September 30, 2021 ageal Hypercontractility in Mice. In control, SCD, and PHZ esophagus with BAY 41-2272 (1 mM) reduced the increased mice, the muscarinic agonist CCh (10 nM to 100 mM) produced neurogenic esophagus contractions in PHZ mice, but contrac- concentration-dependent esophagus contractions (Fig. 2A). tions remained unchanged with the preincubation with SNP The values of Emax were significantly greater in the PHZ mice (300 mM; Fig. 5B). In control mice, preincubation of esophagus

Fig. 1. Red blood cell (A), hemoglobin (B), plasma hemoglobin (C), hematocrit (D), lactate dehydrogenase (E), and reticulocyte (F). Data are shown as the mean 6 S.E.M. of five to eight mice per group. *P , 0.05 vs. control group; †P , 0.05 vs. PHZ group. Effects of Intravascular Hemolysis on Esophagus 197

Fig. 2. Concentration-response curves for CCh (10 nM to100 mM) (A and C) or KCl (1 mM to 1 M) (B and D) in esophageal rings from control, PHZ, and SCD mice. Data are shown as the mean 6 S.E.M. of five mice per group. *P , 0.05 vs. control group. Downloaded from jpet.aspetjournals.org with SNP (300 mM) or BAY 41-2272 (1 mM) did not change the thrombosis, and alterations of vasomotor tone in patients with neurogenic contractions evoked by EFS (Fig. 5C). PNH or SCD (Moyo et al., 2004; Villagra et al., 2007; Hill et al., Esophagus Smooth Muscles from PHZ Mice Display 2010). Moreover, excess plasma hemoglobin has been associ- Increased Oxidative Stress. The protein expression for the ated with esophageal spasm in PNH patients (Moyo et al., 3-NT was about 2-fold higher in PHZ esophagi compared with 2004; Parker et al., 2005). Therefore, considering that in- control mice (P , 0.05; Fig. 6A). Although we noticed a trend travascular hemolysis leads to NO deficiency and alterations

toward an increase in protein expressions for 4-HNE in the of smooth muscle tone, we investigated the effects of in- at ASPET Journals on September 30, 2021 PHZ group (P 5 0.06), this did not reach statistical signifi- travascular hemolysis on esophagus smooth muscle contrac- cance (Fig. 6B). tile mechanisms in two murine models that display different Intravascular Hemolysis Did Not Amplify the Esoph- levels of intravascular hemolysis, SCD and PHZ mice. In fact, ageal Hypercontractility in eNOS2/2 Mice. The musca- PHZ-treated mice displayed a significant increase in plasma rinic agonist CCh (10 nM to 100 mM; Fig. 7A) and KCl (1 mM hemoglobin compared with SCD mice. In our study, we to 1 M; Fig. 7C) caused concentration-dependent esophagus analyzed the in vitro receptor-dependent (muscarinic receptor contractions in control, eNOS2/2, and PHZ-treated eNOS2/2 agonist CCh) and receptor-independent (responses to KCl) mice. The CCh (Fig. 7B) and KCl (Fig. 7D) Emax values contractile responses in esophagi from PHZ and SCD mice. were significantly higher in the esophagi of eNOS2/2 and The maximal contractile response to CCh and KCl was higher eNOS2/2-PHZ mice compared with control mice. Induction of in the PHZ group, but no significant differences were observed intravascular hemolysis did not modify the higher Emax in SCD mice. It is likely that reduction of basal NO by plasma elicited by CCh and KCl in esophagi from eNOS2/2 mice. hemoglobin may favor the hypercontractile state of the No significant changes in pEC50 for CCh was found between esophagus smooth muscle in PHZ mice. On the other hand, groups (5.83 6 0.08, 5.86 6 0.09, and 5.89 6 0.06 for control, our findings indicate that an excess of cell-free hemoglobin is eNOS2/2,andeNOS2/2-PHZ). not abundant enough to promote esophagus injury in SCD. Iron-Chelating Drugs Treatment Did Not Change the Esophageal Hypercontractility in PHZ Mice. In sepa- TABLE 1 rate groups, control and PHZ mice were treated with deferox- Potency (pEC50) values obtained from concentration-response curves in amine (intraperitoneally) or deferiprone (gavage). Treatment esophageal rings from control, PHZ, SCD, eNOS2/2, and eNOS2/2-PHZ with deferoxamine did not modify the CCh- and KCl-induced mice esophagus contractile responses in control and PHZ mice Concentration-response curves for CCh (10 nM to 100 mM) and KCl (1 mM to 1 M) in the absence and in the presence of either SNP (300 mM) or BAY 41-2272 (BAY; 1 mM). (Fig. 8). Similarly, treatment with deferiprone did not modify Data represent the mean 6 S.E.M. of five experiments. the increased CCh Emax in PHZ (0.05 6 0.05 mN/mg; n 5 4) pEC50 and control mice (0.47 6 0.02 mN/mg; n 5 4). Control PHZ SCD CCh 5.78 6 0.05 6.00 6 0.05 5.92 6 0.09 Discussion CCh/SNP 5.81 6 0.08 5.91 6 0.05 n.d. CCH/BAY 5.74 6 0.05 5.88 6 0.07 n.d. In PNH and SCD, elevated plasma hemoglobin levels consume KCl 1.41 6 0.04 1.52 6 0.02 1.15 6 0.12 micromolar concentrations of exogenously supplied NO (Reiter KCl/SNP 1.33 6 0.03 1.41 6 0.05 n.d. 6 6 et al., 2002; Hill et al., 2010). The low NO bioavailability has been KCl/BAY 1.13 0.08 1.43 0.11 n.d. associated with pulmonary hypertension, platelet activation, n.d., not determined. 198 Henrique Silva et al.

Fig. 3. Concentration-response curves for CCh (10 nM to 100 mM) (A and B) or KCl (1 mM to 1 M) (C and D) in esophageal rings from control, PHZ, and SCD mice in the absence or presence of SNP (300 mM). Data are shown as the mean 6 S.E.M. of five mice per group. Downloaded from jpet.aspetjournals.org Previous studies reported that intravenous administration of knowledge, our study is the first to show that eNOS2/2 mice recombinant hemoglobin in healthy human volunteers caused display increased esophagus contractile mechanisms. increases in esophageal peristaltic amplitude, dysphagia, The NO-cGMP signaling pathway is the main mediator of and abdominal pain (Murray et al., 1991; Viele et al., 1997; smooth muscle relaxation of the esophagus (Goyal and Carmichael et al., 2000). Chaudhury, 2008). Several NO donor compounds can supply To reinforce that dysregulation of the NO pathway leads to a exogenous NO to tissues and cells. SNP is an inorganic

hypercontractile state in the esophagus, we evaluated the compound that provides NO to biologic systems by enzymatic at ASPET Journals on September 30, 2021 esophagus contractile mechanisms in eNOS2/2 mice, a model and nonenzymatic mechanisms (Kowaluk et al., 1992; that displays low NO bioavailability. In fact, eNOS2/2 mice Bonaventura et al., 2008). In our study, we tested the effects displayed a hypercontractile state in the esophagus compared of NO supplementation by SNP with the aim of reversing the with control mice. In eNOS2/2 mice, induction of intravascu- esophageal hypercontractility from PHZ mice. In vitro lar hemolysis by PHZ did not modify the esophagus contractile incubation with SNP did not modify increased CCh- and response. Our findings strongly indicate that intravascular KCl-induced contractions in esophagi from PHZ mice. It is hemolysis has a significant function in the development of likely that hemoglobin in PHZ mouse esophagi acts to esophageal dysfunction by reducing NO bioavailability. To our eliminate NO before it binds sGC. A previous study reported

Fig. 4. Concentration-response curves for CCh (10 nM to 100 mM) (A and B) or KCl (1 mM to 1 M) (C and D) in esophageal rings from control, PHZ, and SCD mice in the absence or presence of BAY 41-2272 (1 mM). Data are shown as the mean 6 S.E.M. of five mice per group. †P , 0.05 vs. PHZ group. Effects of Intravascular Hemolysis on Esophagus 199

in vitro incubation with BAY 41-2272 completely restored the responses to CCh and KCl in esophagi from PHZ mice, indicating that cGMP accumulation into esophagus smooth muscle acts to neutralize the NO deficiency. The selective phosphodiesterase 5 inhibitor enhances intracellu- lar cGMP levels and elevates NO-dependent smooth muscle relaxations (Francis et al., 2010). Consistent with our result, previous clinical studies showed that sildenafil decreases esophageal spasm and peristaltic amplitude in patients with esophageal motor alterations (Bortolotti et al., 2002; Eherer et al., 2002; Fox et al., 2007). Neurotransmission is negatively modulated by NO released from nitrergic nerves (Bult et al., 1990). In fact, in vitro incubation of the esophagus body with L-NG-Nitroarginine; NG-nitro-L-Arginine (nonspecific nitric oxide synthase inhib- itor) increases the EFS-induced esophageal contractile re- sponses (Lecea et al., 2011). Previous studies showed that neuronal NO synthase gene-deficient mice display uncoordi- Downloaded from nated and higher esophagus contractions compared with wild mice (Müller et al., 2014). In this study, neurogenic contrac- tions were greater in esophagi from PHZ compared with control mice, suggesting increased nerve-evoked acetylcholine release in the esophagus. This supports that the lack of neuronal nitric oxide synthase-derived NO contributions to jpet.aspetjournals.org esophageal hypercontractility in PHZ mice. Again, in vitro incubation with BAY 41-2272 (1 mM) completely reversed the increased esophagus contractile responses to EFS in PHZ mice, but responses remained unchanged with preincubation with SNP. It is possible that the restoration of contractions by BAY 41-2272 in PHZ mice results in a high cGMP accumula-

tion in the esophageal smooth muscle. at ASPET Journals on September 30, 2021 Oxidative stress has a significant function in the patho- physiology of hemolytic anemia, contributing to organ injury 2 (Fibach and Dana, 2015). Elevated O2 production reacts promptly with NO, reducing NO bioavailability and generat- ing peroxynitrite, which is a potent oxidant and cytotoxic agent (Pacher et al., 2007). Peroxynitrite is a reactive nitrogen species that can cause the nitration of tyrosine residues, generating 3-NT (Pacher et al., 2007). The 3-NT protein is considered a consistent biomarker for evaluating the reactive nitrogen species formation (Halliwell, 1997). 4-HNE, a rela- tively stable end product of lipid peroxidation, is considered a biomarker of oxidative stress (Cohen et al., 2013). In our study, protein expression of 3-NT was increased in PHZ mice compared with the control group. Although there was a trend toward increased protein expression of 4-HNE in PHZ mice, it did not achieve statistical significance. Collectively, our data 2 Fig. 5. Contractile responses to EFS (8–32 Hz) in esophageal rings from suggest that excess O2 reduces NO bioavailability in esophagi control (A and C) and PHZ mice (A and B) in the absence or presence of from PHZ mice. In this circumstance, it is plausible to suggest SNP (300 mM) or BAY 41-2272 (1 mM). Data are shown as the mean 6 2 S.E.M. of five mice per group. *P , 0.05 vs. control group; †P , 0.05 that the enhanced O2 production in esophageal smooth vs. PHZ group. muscle acts to inactivate NO derived from either neuronal nitric oxide synthase or endothelial nitric oxide synthase, as that extravascular translocation of plasma hemoglobin into well as exogenous NO released from SNP. Consistent with this interstitial spaces impairs the vasodilation induced by NO hypothesis, previous studies showed that oxidative stress donor in pig and rat coronary (Schaer et al., 2013). reduces relaxations induced by SNP in mesenteric Recently, sGC stimulators, such as BAY 41-2272, emerged as and erectile tissue (Jiménez-Altayó et al., 2006; Silva et al., potentially better therapeutic drugs because they can stim- 2013). ulate sGC mostly via NO-independent mechanisms (Stasch In hemolytic anemia, accumulation of excess iron can occur and Evgenov, 2013). In circumstances where excess concen- in certain organs due to increased hemoglobin catabolism (Shi tration of plasma hemoglobin is enough to reduce NO bio- et al., 2016). Considering that free iron participates in the availability, sGC stimulators are beneficial to normalize generation of reactive oxygen species (Winterbourn, 1995), we cGMP-dependent vasodilation (Raat et al., 2013). In our study, evaluated the effects of treatment with two iron-chelating 200 Henrique Silva et al.

Fig. 6. Representative images of Western blotting (top panels) and protein values (bottom panels) for 3-NT (A and B, respectively) and 4-HNE (C and D, respectively) in homogenates of esophagi from control (CTL) and PHZ mice. Data are shown as the mean 6 S.E.M. of five mice per group. *P , 0.05 vs. control group. Downloaded from jpet.aspetjournals.org compounds (Pepe et al., 2011), deferiprone and deferoxamine, to oxidative stress, a known complication (Kalogeris et al., on esophageal hypercontractility in PHZ mice. PHZ mouse 2012). In fact, protein expression of 3-NT and 4-HNE in- treatment with deferiprone and deferoxamine did not change creased in esophagi from PHZ mice. Therefore, esophageal hypercontractile state in the esophagus, indicating that excess vascular insufficiency may also play an important role in the of free iron is probably not associated with this disorder. development of an esophageal hypercontractile state. After hemolysis, free hemoglobin can translocate across the In summary, our results show that the hypercontractile vascular wall and reach tissue interstitial spaces (Schaer state seen in esophagi from PHZ mice is associated with et al., 2013). A limitation of our study is that we did not decreased NO bioavailability in esophagus smooth muscle due at ASPET Journals on September 30, 2021 evaluate the localization of hemoglobin and nitric oxide to exaggerated intravascular hemolysis and oxidative stress. synthases in the esophagus. However, previous evidence Moreover, BAY 41-2272 reversed the esophageal hypercon- (Schaer et al., 2016) supports that free hemoglobin can tractility from PHZ mice, indicating that sGC stimulators may scavenge nitric oxide in the vessels and causes vascular be an interesting approach to treat esophageal dysfunction in insufficiency in the esophageal vascular bed, impairing blood conditions of oxidative stress and elevated plasma hemoglobin perfusion and resulting in ischemia. Ischemic conditions lead levels. SCD mice do not present with obvious esophageal

Fig. 7. Concentration-response curves for CCh (10 nM to 100 mM) (A) or KCl (1 mM to 1 M) (C) in esophageal rings from control, eNOS2/2, and eNOS2/2-PHZ mice. The maximal response values are shown in panels B and D. Data are shown as the mean 6 S.E.M. of five mice per group. *P , 0.05 vs. control group. Effects of Intravascular Hemolysis on Esophagus 201

Fig. 8. Concentration-response curves for CCh (10 nM to 100 mM) (A and B) or KCl (1 mM to 1 M) (C and D) in esophageal rings from control and PHZ mice treated or not treated with deferoxamine. Data are shown as the mean 6 S.E.M. of five experiments per group. *P , 0.05 vs. control group. Downloaded from jpet.aspetjournals.org disorder, indicating that the amount of excessive free hemo- Dutra FF, Alves LS, Rodrigues D, Fernandez PL, de Oliveira RB, Golenbock DT, Zamboni DS, and Bozza MT (2014) Hemolysis-induced lethality involves inflam- globin is insufficient to cause esophageal damage in SCD. To masome activation by heme. Proc Natl Acad Sci USA 111:E4110–E4118. the best of our knowledge, this study provides the first Eherer AJ, Schwetz I, Hammer HF, Petnehazy T, Scheidl SJ, Weber K, and Krejs GJ (2002) Effect of sildenafil on oesophageal motor function in healthy subjects and experimental evidence to support the hypothesis that esoph- patients with oesophageal motor disorders. Gut 50:758–764. ageal symptoms in PNH patients are secondary to intravas- Fibach E and Dana M (2015) Oxidative stress in paroxysmal nocturnal hemoglo- binuria and other conditions of complement-mediated hemolysis. Free Radic Biol cular hemolysis affecting the NO-cGMP pathway. –

Med 88 (Pt A):63 69. at ASPET Journals on September 30, 2021 Förstermann U and Sessa WC (2012) Nitric oxide synthases: regulation and function. Acknowledgments Eur J 33:829–837, 837a-837d. Fox M, Sweis R, Wong T, and Anggiansah A (2007) Sildenafil relieves symptoms and We are thankful for the access to equipment and assistance normalizes in patients with oesophageal spasm: a report of two cases. provided by Dr. Edson Antunes’ laboratory (Department of Pharma- Neurogastroenterol Motil 19:798–803. cology, University of Campinas). Francis SH, Busch JL, Corbin JD, and Sibley D (2010) cGMP-dependent protein kinases and cGMP phosphodiesterases in nitric oxide and cGMP action. Pharmacol Rev 62:525–563. Authorship Contributions Goyal RK and Chaudhury A (2008) Physiology of normal esophageal motility. J Clin Participated in research design: Henrique Silva, Yotsumoto Fertrin, Gastroenterol 42:610–619. Groneberg D, Zizer E, Lies B, Seidler B, Saur D, Wagner M, and Friebe A (2015) Ferreira Costa. Dominant role of interstitial cells of Cajal in nitrergic relaxation of murine lower Conducted experiments: Henrique Silva, Fernanda Franco-Penteado, oesophageal sphincter. J Physiol 593:403–414. Alexandre, Calmasini. Halliwell B (1997) What nitrates tyrosine? Is nitrotyrosine specific as a biomarker of peroxynitrite formation in vivo? FEBS Lett 411:157–160. Contributed new reagents or analytic tools: Ferreira Costa. Hill A, DeZern AE, Kinoshita T, and Brodsky RA (2017) Paroxysmal nocturnal Performed data analysis: Henrique Silva, Ferreira Costa. haemoglobinuria. Nat Rev Dis Primers 3:17028. Wrote or contributed to the writing of the manuscript: Henrique Hill A, Rother RP, Wang X, Morris SM, Jr, Quinn-Senger K, Kelly R, Richards SJ, Bessler M, Bell L, Hillmen P, et al. (2010) Effect of eculizumab on haemolysis- Silva, Yotsumoto Fertrin, Ferreira Costa. associated nitric oxide depletion, dyspnoea, and measures of pulmonary hyper- tension in patients with paroxysmal nocturnal haemoglobinuria. Br J Haematol References 149:414–425. Bonaventura D, Lunardi CN, Rodrigues GJ, Neto MA, and Bendhack LM (2008) A Inamori M, Shimamura T, Nagase H, Abe Y, Umezawa T, Nakajima A, Saito T, Ueno novel mechanism of vascular relaxation induced by sodium nitroprusside in the N, Tanaka K, Sekihara H, et al. (2006) mRNA expression of inducible nitric oxide isolated rat aorta. Nitric Oxide 18:287–295. synthase, endothelial nitric oxide synthase and vascular endothelial growth factor Bortolotti M, Pandolfo N, Giovannini M, Mari C, and Miglioli M (2002) Effect of in esophageal mucosa biopsy specimens from patients with reflux . Sildenafil on hypertensive lower oesophageal sphincter. Eur J Clin Invest 32: Hepatogastroenterology 53:361–365. 682–685. Jiménez-Altayó F, Briones AM, Giraldo J, Planas AM, Salaices M, and Vila E (2006) Braverman AS, Miller LS, Vegesna AK, Tiwana MI, Tallarida RJ, and Ruggieri MR, Increased superoxide anion production by interleukin-1beta impairs nitric oxide- Sr (2009) Quantitation of the contractile response mediated by two receptors: M2 mediated relaxation in resistance arteries. J Pharmacol Exp Ther 316:42–52. and M3 muscarinic receptor-mediated contractions of human gastroesophageal Kalogeris T, Baines CP, Krenz M, and Korthuis RJ (2012) Cell biology of ischemia/ smooth muscle. J Pharmacol Exp Ther 329:218–224. reperfusion injury. Int Rev Cell Mol Biol 298:229–317. Bult H, Boeckxstaens GE, Pelckmans PA, Jordaens FH, Van Maercke YM, Kato GJ, Steinberg MH, and Gladwin MT (2017) Intravascular hemolysis and the and Herman AG (1990) Nitric oxide as an inhibitory non-adrenergic non- pathophysiology of sickle cell disease. J Clin Invest 127:750–760. cholinergic neurotransmitter. Nature 345:346–347. Kowaluk EA, Seth P, and Fung HL (1992) Metabolic activation of sodium nitroprusside Bunn HF, Nathan DG, Dover GJ, Hebbel RP, Platt OS, Rosse WF, and Ware RE to nitric oxide in vascular smooth muscle. JPharmacolExpTher262:916–922. (2010) Pulmonary hypertension and nitric oxide depletion in sickle cell disease. Lecea B, Gallego D, Farré R, Opazo A, Aulí M, Jiménez M, and Clavé P (2011) Blood 116:687–692. Regional functional specialization and inhibitory nitrergic and nonnitrergic con- Carmichael FJL, Ali ACY, Campbell JA, Langlois SF, Biro GP, Willan AR, Pierce CH, eurotransmission in the human esophagus. Am J Physiol Gastrointest and Greenburg AG (2000) A phase I study of oxidized raffinose cross-linked human Physiol 300:G782–G794. hemoglobin. Crit Care Med 28:2283–2292. Lim SK, Kim H, Lim SK, bin Ali A, Lim YK, Wang Y, Chong SM, Costantini F, Cohen G, Riahi Y, Sunda V, Deplano S, Chatgilialoglu C, Ferreri C, Kaiser N, and Baumman H (1998) Increased susceptibility in Hp knockout mice during acute and Sasson S (2013) Signaling properties of 4-hydroxyalkenals formed by lipid hemolysis. Blood 92:1870–1877. peroxidation in diabetes. Free Radic Biol Med 65:978–987. Mearin F, Mourelle M, Guarner F, Salas A, Riveros-Moreno V, Moncada S, Crosby WH (1955) The metabolism of hemoglobin and bile pigment in hemolytic and Malagelada JR (1993) Patients with achalasia lack nitric oxide synthase disease. Am J Med 18:112–122. in the gastro-oesophageal junction. Eur J Clin Invest 23:724–728. 202 Henrique Silva et al.

Moyo VM, Mukhina GL, Garrett ES, and Brodsky RA (2004) Natural history of Shah V, Lyford G, Gores G, and Farrugia G (2004) Nitric oxide in gastrointestinal paroxysmal nocturnal haemoglobinuria using modern diagnostic assays. Br J health and disease. 126:903–913. Haematol 126:133–138. Shi PA, Choi E, Chintagari NR, Nguyen J, Guo X, Yazdanbakhsh K, Mohandas N, Müller M, Colcuc S, Drescher DG, Eckardt AJ, von Pein H, Taube C, Schumacher J, Alayash AI, Manci EA, Belcher JD, et al. (2016) Sustained treatment of sickle cell Gockel HR, Schimanski CC, Lang H, et al. (2014) Murine genetic deficiency of mice with haptoglobin increases HO-1 and H-ferritin expression and decreases iron neuronal nitric oxide synthase (nNOS(-/-)) and interstitial cells of Cajal (W/W(v)): deposition in the without improvement in kidney function. Br J Haematol implications for achalasia? J Gastroenterol Hepatol 29:1800–1807. 175:714–723. Murray J, Du C, Ledlow A, Bates JN, and Conklin JL (1991) Nitric oxide: mediator of Silva FH, Mónica FZ, Báu FR, Brugnerotto AF, Priviero FBM, Toque HA, nonadrenergic noncholinergic responses of opossum esophageal muscle. Am J and Antunes E (2013) Superoxide anion production by NADPH oxidase plays a Physiol 261:G401–G406. major role in erectile dysfunction in middle-aged rats: prevention by antioxidant Oni SB, Osunkoya BO, and Luzzatto L (1970) Paroxysmal nocturnal hemoglobinuria: therapy. J Sex Med 10:960–971. evidence for monoclonal origin of abnormal red cells. Blood 36:145–152. Stasch J-P and Evgenov OV (2013) Soluble guanylate cyclase stimulators in pul- Pacher P, Beckman JS, and Liaudet L (2007) Nitric oxide and peroxynitrite in health monary hypertension. Handb Exp Pharmacol 218:279–313. and disease. Physiol Rev 87:315–424. Viele MK, Weiskopf RB, and Fisher D (1997) Recombinant human hemoglobin does Parker C, Omine M, Richards S, Nishimura J, Bessler M, Ware R, Hillmen P, Luzzatto not affect renal function in humans: analysis of safety and pharmacokinetics. L, Young N, Kinoshita T, et al.; International PNH Interest Group (2005) Diagnosis Anesthesiology 86:848–858. and management of paroxysmal nocturnal hemoglobinuria. Blood 106:3699–3709. Villagra J, Shiva S, Hunter LA, Machado RF, Gladwin MT, and Kato GJ (2007) Pepe A, Meloni A, Capra M, Cianciulli P, Prossomariti L, Malaventura C, Putti MC, Platelet activation in patients with sickle disease, hemolysis-associated pulmonary Lippi A, Romeo MA, Bisconte MG, et al. (2011) Deferasirox, deferiprone and des- hypertension, and nitric oxide scavenging by cell-free hemoglobin. Blood 110: ferrioxamine treatment in thalassemia major patients: cardiac iron and function 2166–2172. comparison determined by quantitative magnetic resonance imaging. Haemato- Winterbourn CC (1995) Toxicity of iron and hydrogen peroxide: the Fenton reaction. logica 96:41–47. Toxicol Lett 82–83:969–974. Raat NJH, Tabima DM, Specht PAC, Tejero J, Champion HC, Kim-Shapiro DB, Yamato S, Spechler SJ, and Goyal RK (1992) Role of nitric oxide in esophageal Baust J, Mik EG, Hildesheim M, Stasch J-P, et al. (2013) Direct sGC activation peristalsis in the opossum. Gastroenterology 103:197–204. bypasses NO scavenging reactions of intravascular free oxy-hemoglobin and limits Yazaki E and Sifrim D (2012) Anatomy and physiology of the esophageal body. Dis vasoconstriction. Antioxid Redox Signal 19:2232–2243. Esophagus 25:292–298. Downloaded from Reiter CD, Wang X, Tanus-Santos JE, Hogg N, Cannon RO, III, Schechter AN, Zhang L, Prak L, Rayon-Estrada V, Thiru P, Flygare J, Lim B, and Lodish HF (2013) and Gladwin MT (2002) Cell-free hemoglobin limits nitric oxide bioavailability in ZFP36L2 is required for self-renewal of early burst-forming unit erythroid pro- sickle-cell disease. Nat Med 8:1383–1389. genitors. Nature 499:92–96. Schaer CA, Deuel JW, Schildknecht D, Mahmoudi L, Garcia-Rubio I, Owczarek C, Schauer S, Kissner R, Banerjee U, Palmer AF, et al. (2016) Haptoglobin preserves vascular nitric oxide signaling during hemolysis. Am J Respir Crit Care Med 193: Address correspondence to: Dr. Fabio Henrique Silva, Hematology and 1111–1122. Hemotherapy Center, University of Campinas-UNICAMP, Rua Carlos Chagas, Schaer DJ, Buehler PW, Alayash AI, Belcher JD, and Vercellotti GM (2013) Hemo- 480, Cidade Universitária, 13083-970, Campinas-SP, Brazil. E-mail: fabioh- jpet.aspetjournals.org lysis and free hemoglobin revisited: exploring hemoglobin and hemin scavengers as [email protected] a novel class of therapeutic proteins. Blood 121:1276–1284. at ASPET Journals on September 30, 2021